Egger's Test for Publication Bias

Egger's test for publication bias is a regression test for funnel plot asymmetry in meta-analysis, regressing standardised effect sizes on their precision and testing whether the intercept differs from zero.

에거 검정은 표준화한 효과크기를 정밀도에 회귀시킨 뒤 절편이 0과 다른지를 검정하여 메타분석의 깔때기 그림 비대칭을 확인하는 회귀 기반 출판편향 검정이다.

How do I test whether my meta-analysis shows signs of publication bias or small-study effects?

메타분석에 출판편향이나 소규모 연구 효과가 있는지 어떻게 검정하나요?

Egger's test regresses each study's effect size divided by its standard error on the inverse of that standard error, and an intercept that differs from zero indicates funnel plot asymmetry. Asymmetry can arise from publication bias but also from genuine heterogeneity or other small-study effects, and the test has low power with fewer than about 10 studies.

When to use Egger's Test for Publication Bias

• A meta-analysis of continuous effect sizes with at least about 10 studies
• Testing funnel plot asymmetry formally rather than only by eye
• Reporting small-study effects alongside the pooled estimate

When to use something else

• Log odds ratios from binary outcomes, where Egger's test has an inflated false-positive rate: use Peters' or Harbord's test
• Fewer than about 10 studies: rely on a funnel plot and sensitivity analyses such as trim-and-fill, interpreted cautiously

Assumptions and requirements

• One effect size and its sampling variance per study (the variance, not the standard error, is entered)
• Studies independent of one another
• Asymmetry not driven mainly by between-study heterogeneity

What the analysis reports

• Intercept (bias estimate) with standard error
• t statistic and two-sided p-value for the intercept
• Warning when fewer than 10 studies are included

How to report it

Egger's regression test indicated no significant funnel plot asymmetry, intercept = [value], SE = [value], t([df]) = [value], p = [value].

Related methods

• Meta Analysis
• Meta Regression

Run Egger's Test for Publication Bias on Quorant